\documentclass[11pt]{article}

\usepackage{natbib}
\usepackage{etex}
\usepackage[flushleft]{threeparttable} 
\usepackage[utf8]{inputenc}
\usepackage{amsmath}
\usepackage{amssymb,amsfonts,textcomp}
\usepackage{array}
\usepackage{multirow} 
\usepackage{hhline}
\usepackage[usenames,dvipsnames]{color}
\usepackage{rotating}
\usepackage{adjustbox}
\usepackage{enumerate}
\usepackage{url}
\usepackage{booktabs}
\usepackage{epstopdf}
\usepackage[small, sc, textfont={sc}, justification=raggedright, hangindent=5.7em]{caption}
\usepackage{graphicx}
\usepackage[capposition=top]{floatrow}
\usepackage{dcolumn}
\usepackage{bm}
\usepackage[sc]{mathpazo}
\usepackage[T1]{fontenc}
\usepackage[margin=1in]{geometry}
\usepackage[british,UKenglish,USenglish,english,american]{babel}
\usepackage{microtype}										
\usepackage{booktabs}
\usepackage{setspace}
\usepackage{ulem}
\usepackage{threeparttable}
\usepackage[usenames,dvipsnames]{color}
\usepackage{xr-hyper}
\usepackage{pdflscape}
\usepackage{bbm}
\usepackage[pdfborder={0 0 0}, bookmarksnumbered=true, bookmarksopenlevel=1, colorlinks=true, allcolors=BrickRed, citecolor=Blue, hyperfootnotes=false]{hyperref}
\usepackage{titlesec}
\usepackage{subcaption}

\begin{document}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Descriptive Statistics
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%


\begin{table}[htbp]
\begin{threeparttable}
\caption{Descriptive Statistics of Working Children}
\label{tab:dessectors}
\begin{tabular}{lcccccc}
\toprule
&\multirow{2}{7ex}{\centering All} &\multirow{2}{7ex}{\centering Boys} & \multirow{2}{7ex}{\centering Girls}  & \multicolumn{3}{c}{Children Aged [\dots]}\\ \cmidrule{5-7}
&&&&12-14& 15-16 & 17-18\\ 
& (1) & (2)  & (3) & (4) & (5) & (6) \\
\midrule
\\[-1em]
\textit{Economic Activity}&\\
\input{wholesale}
\input{hotels}
\input{manufacture}
\input{transportation}
\input{other}
\input{construction}
\input{domestic}
\input{others}
\midrule 
\input{averagehours.tex}
\input{work_study}
\input{averagework_home}
\bottomrule
\end{tabular}
		\begin{tablenotes}[flushleft]
\setlength\labelsep{0pt}			\item\small \emph{Notes:} The sample includes children between 12-18 years old from \emph{GEIH, 2008-2010.} We use the 2-digit International Standard Industrial Classification (ISIC) to classify  the economic activities. We present the five most important economic sectors, in addition to the construction and domestic services sector. None of the omitted activities (aggregated as ``Others'') accounts for more than five percent of working children.
		\end{tablenotes}
	\end{threeparttable}
\end{table}

\begin{table}[htbp]
\begin{threeparttable}
\caption{Descriptive Statistics}
\label{tab:remit}
\begin{tabular}{lcc}
\toprule
 & Recipient & Non-recipient\\ 
&(1)& (2) \\
\midrule
\textit{Outcome variables}&&\\
\input{averageoutcomes.tex} [.5em]
\input{averageoutcomeshours.tex} [.5em]
\input{averagehouseworkshares.tex} [.5em]
\textit{Children characteristics} & & \\ 
\input{averagechildchars.tex}  [.5em]
\textit{Household head characteristics}  & &\\ 
\input{averagehouseholdheadchars.tex} [.5em] 
\textit{Household characteristics}  & &\\ 
\input{averagehouseholdchars.tex} [.5em]
\textit{Additional statistics} \\
\multirow{2}{28ex}{Share in bottom metro-area income quartile} & 0.044 & 0.304 \\ 
\\ [0.5em]
\hline
Observations & 4013 & 139642\\ 
\bottomrule
\end{tabular}
		\begin{tablenotes}[flushleft]
\setlength\labelsep{0pt}			\item\small \emph{Notes:} The sample includes children between 12-18 years old from \emph{GEIH, 2008-2010}. Total amount of remittances in PPP-adjusted US dollar.
		\end{tablenotes}
	\end{threeparttable}
\end{table}

\begin{table}[htbp!]
\begin{threeparttable}
\caption{Impact of Unemployment Shocks on Remittances}
\label{tab:remitshock}
\begin{tabular}{lccc}
\toprule
\input{remittances_shock.tex}
\midrule
Region FE &Yes & Yes & Yes \\
Child characteristics & Yes & Yes & Yes \\
Household characteristics & Yes & Yes & Yes \\
Region Variables & Yes & Yes  & Yes \\
\bottomrule
\end{tabular}
\begin{tablenotes}[flushleft]
\setlength\labelsep{0pt} \item\small \emph{Notes:} The sample includes children between 12-18 years old from \emph{GEIH, 2008-2010}. The individual and household controls included are presented in Table 2. Regional controls include the Gini coefficient, the unemployment rate, and the regional GDP growth. Coefficients are reported with standard errors clustered at the region-month-year level in parentheses. */**/*** denotes significance at the 10/5/1 percent level. 
\end{tablenotes} 
\end{threeparttable}
\end{table}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% MAIN RESULTS
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

\begin{table}[htbp]
	\begin{adjustbox}{max width=\textwidth}
	\begin{threeparttable}
	\caption{Impact of Remittances on Children Outcomes}
	\label{tab:mainEst}
	\begin{tabular}{lcccccc}
	\toprule
	\multirow{2}{30ex}{Dependent variable\\ (Panels A and B):}&\multicolumn{2}{c}{Likelihood to work} &\multicolumn{2}{c}{Hours} &\multicolumn{2}{c}{School attendance} \\
	&\multicolumn{2}{c}{(Linear probability model)} &\multicolumn{2}{c}{(Tobit)} &\multicolumn{2}{c}{(Linear probability model)} \\
		 \cmidrule(lr){2-3}  \cmidrule(lr){4-5} \cmidrule(lr){6-7}
		& (1) & (2)  & (3) & (4) & (5) & (6) \\
	\hline 
	\\ [-0.3em]
	\emph{Panel A: Non-IV Estimates} \\
	\input{childrenNonIV}      
	\\ 
	\emph{Panel B: IV Estimates}  \\  
	\input{children_IV}                                   
	\\ 
	\\ [-0.75em]
	Dependent variable (panel C):  &\multicolumn{6}{c}{Remittances}\\    
	\hline                                 
	\\ [-0.75em]
	\emph{Panel C: First Stage}\\
	\input{childrenfstage}
	\\ [-0.75em] 
	\hline 
	Region  FE & Yes & Yes & Yes & Yes & Yes & Yes\\
	Child and HH characteristics & No & Yes & No & Yes & No & Yes \\
	Region Variables & No& Yes & No & Yes& No & Yes \\
	\hline
	Observations   & 143655   &   143655   &  143655 & 143655 & 143655 & 143655   \\
	\bottomrule
	\end{tabular}
			\begin{tablenotes}[flushleft]
	\setlength\labelsep{0pt}			\item\small \emph{Notes:} The sample includes children between 12-18 years old from \emph{GEIH, 2008-2010}. The individual and household controls included are presented in Table~\ref{tab:remit}. Regional controls include the Gini coefficient, the unemployment rate, and the regional GDP growth. Columns 1-2 and 5-6 in Panel B report coefficients from the LPM with standard errors clustered at the region-month-year level in parentheses. Columns 4-6 in Panel B report the marginal effect from the IV-Tobit calculated based on Wooldridge(2010, Section 17.2). The standard error is bootstrapped with 500 replications and clustered at the region-month-year level. The point estimates and standard errors for the LPM, and the marginal effect and standard error for the Tobit model in Panel A and Panel B are multiplied by 100 for presentation purposes. */**/*** denotes significance at the 10/5/1 percent level. 
			\end{tablenotes}	
			\end{threeparttable}
	\end{adjustbox}
\end{table}

\begin{table}[htbp!]
\begin{adjustbox}{max width=\textwidth}
\begin{threeparttable}
\caption{Robustness Checks on Export-Income Share and Return Migration}
\label{tab:threats}
\begin{tabular}{lcccc}
\toprule
&\multicolumn{4}{c}{Likelihood to work} \\
 \cmidrule(lr){2-5}
& (1) & (2)  & (3) & (4) \\
\hline
\input{children_threats}                                                
\bottomrule
\end{tabular}
		\begin{tablenotes}[flushleft]
\setlength\labelsep{0pt}\item\small \emph{Notes:} The sample includes children between 12-18 years old from \emph{GEIH, 2008-2010}. The individual and household controls included are presented in Table~2. Regional controls include the Gini coefficient, the unemployment rate, and the regional GDP growth. Coefficients are reported with standard errors clustered at the region-month-year level in parentheses. The point estimates and standard errors are multiplied by 100 for presentation purposes. */**/*** denotes significance at the 10/5/1 percent level.
		\end{tablenotes}
	\end{threeparttable}
\end{adjustbox}	
\end{table}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% HETEROGENEUS RESULTS
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

\begin{table}[htbp!]
\begin{adjustbox}{max width=\textwidth}
\begin{threeparttable}
\caption{Heterogeneous Treatment Effect of Remittances By Age and Gender}
\label{tab:hetero}
\begin{tabular}{lcccc}
\toprule
 & \multicolumn{2}{c}{Children Aged [\dots]} &\multirow{2}{7ex}{\centering Boys} & \multirow{2}{7ex}{\centering Girls} \\ \cmidrule{2-3}
&12-14& 15-18\\ 
  & (1) & (2)  & (3) & (4)  \\
\midrule
\\ [-0.5em]
&\multicolumn{4}{c}{\emph{Panel A. Dependent variable: Labor market participation}}   \\ [0.15em] \cmidrule{2-5}
\input{heteroLPMPanelA.tex}           
\hline
\input{averageheteroPanelA.tex}
\bottomrule
\\ [-0.75em]
& \multicolumn{2}{c}{Children Aged 12-14} &\multicolumn{2}{c}{Children Aged 15-18} \\
 \cmidrule(lr){2-3}  \cmidrule(lr){4-5} 
 & Boys & Girls & Boys & Girls \\ 
  & (5) & (6)  & (7) & (8)  \\
\midrule
\\ [-0.5em]
&\multicolumn{4}{c}{\emph{Panel B1. Dependent variable: Labor market participation}}   \\ [0.15em] \cmidrule{2-5}
\input{heteroLPMPanelB.tex}            
\input{averageheteroPanelB.tex}
\\ [-0.5em]
&\multicolumn{4}{c}{\emph{Panel B2. Dependent variable: Household-work participation}}   \\ [0.15em] \cmidrule{2-5}
\input{housework_hetero.tex}            
\input{averageheteroPanelC.tex}
\\ [-0.75em]
\hline
\\ [-0.75em]         
\bottomrule
\end{tabular}
		\begin{tablenotes}[flushleft]
\setlength\labelsep{0pt}\item\small \emph{Notes:} The sample includes children between 12-18 years old from \emph{GEIH, 2008-2010}. The individual and household controls included are presented in Table~2. Regional controls include the Gini coefficient, the unemployment rate, and the regional GDP growth. Coefficients are reported with standard errors clustered at the region-month-year level in parentheses. The point estimates and standard errors are multiplied by 100 for presentation purposes. */**/*** denotes significance at the 10/5/1 percent level. 
		\end{tablenotes}
	\end{threeparttable}
\end{adjustbox}	
\end{table}

\begin{landscape}
\begin{table}[htbp]
	\begin{adjustbox}{max width=\textwidth}
	\begin{threeparttable}
	\caption{Heterogeneous Treatment Effect of Remittances by Wealth and Age}
	\label{tab:heterowealth}
	\begin{tabular}{lcccccc}
	\toprule
	&\multicolumn{2}{c}{All} &\multicolumn{2}{c}{Children Aged 12-14}  &\multicolumn{2}{c}{Children Aged 15-18}   \\ 
		\cmidrule(lr){2-3} \cmidrule(lr){4-5} \cmidrule(lr){6-7}

		& Poor &Non-poor & Poor &Non-poor & Poor &Non-poor \\ 
	\input{wealthLPM.tex}    
	\hline
	\input{averagewealth.tex}         
	\bottomrule
	\end{tabular}
			\begin{tablenotes}[flushleft]
\setlength\labelsep{0pt}\item\small \emph{Notes:} The sample includes children between 12-18 years old from \emph{GEIH, 2008-2010}. The individual and household controls included are presented in Table~2. Regional controls include the Gini coefficient, the unemployment rate, and the regional GDP growth. Coefficients are reported with standard errors clustered at the region-month-year level in parentheses. The point estimates and standard errors are multiplied by 100 for presentation purposes. */**/*** denotes significance at the 10/5/1 percent level. 
			\end{tablenotes}
		\end{threeparttable}
	\end{adjustbox}	
\end{table}

\begin{table}[htbp!]
	\begin{adjustbox}{max width=\textwidth}
	\begin{threeparttable}
	\caption{Remittances and Household Labor Supply}
	\label{tab:adults}
	\begin{tabular}{lcccccc}
	\toprule
	&\multirow{3}{15ex}{\centering Children} &\multicolumn{5}{c}{Adults (Children-Linked Sample)}\\
	\cmidrule(lr){3-7}
	& &All & Male & Female & Poor & Non-Poor \\ 
	\cmidrule(lr){3-7}                             
	\input{adultshetero.tex}
	\hline
	\input{averageadultshetero.tex}
	\bottomrule
	\end{tabular}
			\begin{tablenotes}[flushleft]
	\setlength\labelsep{0pt}			\item\small \emph{Notes:} The sample includes children between 12-18 years old, and adults between 19-65 years old in households with children from \emph{GEIH, 2008-2010}. The individual and household controls included are presented in Table~\ref{tab:remit}. Regional controls include the Gini coefficient, the unemployment rate, and the regional GDP growth.  Columns report coefficients from the LPM with standard errors clustered at the region-month-year level in parentheses. The point estimates and standard errors are multiplied by 100 for presentation purposes. */**/*** denotes significance at the 10/5/1 percent level. 
			\end{tablenotes}	
			\end{threeparttable}
	\end{adjustbox}
\end{table}

\end{landscape}

\end{document}

